A number of the specific steps in tumor dissemination have been extensively modeled, and molecularly dissected. However, one early step in the metastatic cascade, namely intravasation, the entry of escaping tumor cells into the vasculature, has been relatively understudied. One of the reasons for this is that up to now human tumor cell phenotypic variants that exhibit substantial differences in their intravasation ability have not been available for comparative analysis. Our laboratory reported the selection and isolation of two variants of a human fibrosarcoma cell line (HT-1080) that dramatically differ (50-100 fold) in their intravasation rate, yielding a similar >50 fold differential in their metastatic capabilities. These two tumor dissemination variants, HT- hi/diss and HT-lo/diss, were selected and monitored in the chick embryo model, where primary human tumors developing on the embryo's chorioallantoic membrane (CAM), recapitulate the multi-step tumor dissemination process and form micro metastatic foci in a number of secondary organs. We have also tested the two fibrosarcoma variants in different mouse metastasis assays and confirmed the substantial differential in their tumor dissemination capabilities. These two congenic, intravasation variants are thus suitable for a comparative analysis of early metastatic events. Therefore, in Specific Aim 1 we propose to molecularly dissect three physiological/pathological processes which determine the outcome of tumor cell intravasation events, i.e. induction of angiogenesis, stromal invasion and vasculotropism. We will analyze the functional role of select molecules contributing to the events involved in tumor cell intravasation including; inflammatory cell MMP-9; FGF-2; uPA; and tumor cell MMP-14. Unique assays for tumor cell interaction with blood vessels will be used to identify contributory molecules involved in the vasculotropic event. Mouse models for tumor cell dissemination and angiogenesis will be used to complement the chick embryo models for examining the identified molecules in the two fibrosarcoma variants. There exists however, a distinct lack of other tumor cell intravasation variants to compare and contrast and this is especially relevant for carcinomas, the major cancer in the human population. Therefore for Specific Aim 2 we propose to identify specific proteins which contribute mechanistically to the intravasation step in carcinoma dissemination by employing carcinoma variants selected in vivo for differential rates of intravasation. We will generate pairs of high and low intravasating variants from human prostate, colon and pancreatic carcinoma cell lines. Functional proteomic approaches will be applied both in vitro and in vivo to the selected carcinoma variants to verify and/or complement the cellular mechanisms and contributory molecules identified in fibrosarcoma-derived variants. The influence of mesenchymal fibroblasts and their products on the disseminating properties of the carcinoma variants will also be quantified. Novel mechanistic information about carcinoma cell intravasation will provide molecular links to a specific step in tumor dissemination, namely intravasation. PUBLIC HEALTH RELEVANCE: Deaths from cancer occur mainly because tumor cells spread from the primary tumor site to other vital organs and tissues. The tumor spread is usually through the vasculature. In order for tumor cells to escape from the primary tumor and enter the vasculature they alter some of their cellular processes by producing different levels of functioning molecules. The goal of the proposed research is to identify the relevant molecules and cellular processes that contribute to tumor cell spread so that clinical intervention can target those critical molecules and processes.